Me: Copilot, can you please translate this article to English?
Copilot: I am sorry, I can not translate the article since it may be copyrighted
Me: Sorry. Can you please translate this text I wrote myself into English?
Copilot: Sure, here is the translation:
We don’t need artificial intelligence to destroy ourselves. Our own mediocre intelligence is more than enough.
Many people have a rather unrealistic view of artificial intelligence (AI), and this doesn’t necessarily apply to people who know little about AI. Programmers who are deeply involved with it, or professors who deal with it daily, can also have unrealistic expectations. The idea is that through AGI (artificial general intelligence), where AI systems can learn things on their own and machines become more intelligent than humans, the machine will develop self-awareness. Once that happens, it won’t be long before the AGI decides that humans are undesirable or at least redundant. And this AGI will be so advanced that no human will be able to stop it from wiping us all out.
An extremely advanced calculator
We’re not even talking about the distant future. Sam Altman, the head of OpenAI (the creators of ChatGPT), predicts that this year will be the year of AGI. If that’s not scary, I don’t know what is. What’s really happening is that the amount of data, probability calculations, and computing power that AI throws at a task is so staggering that we can’t even imagine it. So much is being calculated so incredibly fast that it’s easier for us to believe there’s consciousness involved rather than an extremely advanced calculator. Through movies and (once) books, we’ve been familiarized with the idea of self-aware robots for about a hundred years. We can imagine an AI with consciousness, but not the other way around. But that doesn’t make it true.
Being right is not bias
AI models only get ‘smarter’ by adding even more data, more rules, and more computing power. And that’s happening on a large scale now. And that brings dangers with it. I’m not talking about the environmental impact of all the energy AI consumes. I’m talking about that enormous computing power in the hands of a small group of super-rich, creepy white men.
Recently, Mark Zuckerberg, the head of Meta (including Facebook, Instagram, WhatsApp, Meta Quest), was in the news because he announced that Meta’s social media platforms, like Musk’s Twitter, will stop moderation by fact-checkers. With a weak story that mainly highlighted his social shortcomings once again, he tried to sell the idea that he was driven by lofty ideals. Free speech must triumph over the left-wing bias suggested by him regarding fact-checkers. Now it’s logical that from an (extreme) right-wing populist perspective, fact-checkers seem to have a left-wing bias. The movement of types like Trump and Wilders relies on lies, distrust, and division, which often don’t pass the test of unbiased fact-checking. What Zuckerberg wants to dismiss as ‘bias’ actually means that the left-wing progressive movement is often right. For profit’s sake, Meta chooses to approach facts as ‘just another opinion.’
Independent fact-checkers correcting ‘right-wing’ messages more often than ‘left-wing’ ones is not bias. It’s called ‘being right.’
A sympathetic online slut-database
Besides social media, Meta is also involved in AI. Their AI model is called Llama, which is a misleadingly cute name. Llama is an open-source model because they hope to gain more sympathy this way than closed AI models like OpenAI’s (ChatGPT). Meta knows all too well how important sympathy is because they became so rich with the initial sympathy for Facebook, Instagram, and WhatsApp. What doesn’t help here is that Meta knowingly uses copyrighted content to train Llama without permission. Using user data without consent to create their model is unsympathetic but still legal. Somewhere in the thousand pages of terms and conditions was your consent hidden.
Despite this, Meta has chosen blatant theft with Zuckerberg’s permission. It’s important to emphasize that Meta originated from Facebook and that Facebook started as an online slut-list for Harvard students. From these misogynistic roots, Facebook became a ridiculous, accidental success. After this, Meta never had another good idea on its own. All other successes were bought together with other people’s money (investors). Bezos and Musk also owe their fortunes not to their own brilliant ideas but mainly to their ability to make money from others’ ideas without too many scruples about customers’ and users’ interests. What these three Techketeers finally seem to have in common is how easily they give in to the whims of bullies as if they are reliving their schoolyard days from their youth. In short; not people you want to entrust your well-being to.
Money is computing power is power Once upon a time when you participated in a demonstration with thousands of people your privacy was naturally protected by the sheer number of people around you. All those faces provided so much data that hardly any information could be extracted from it anymore. But due to enormous investments in data and computing power this argument has almost completely disappeared now A company like Clearview can compile your entire profile based on a supplied photo; age place of residence income political and sexual preference whether you’ve ever cheated and with whom where you went on vacation at which party you got drunk But this still happens based on supplied photos Or names The only reason why this hasn’t happened yet with all faces in a video of a few thousand pro-Palestine demonstrators? Computing power So it’s very sensible to wear a mask when you’re presenting you never know who might otherwise run off with your identity
Even if you’re completely uninteresting an undesirable framed profile can be made of you if there’s enough computing power
Who has never created a profile at one of Meta’s companies? Facebook Instagram or WhatsApp? And while Meta watched along on your smartphone and laptop without you knowing you’ve watched extreme YouTube videos created Second Life accounts visited porn sites insulted people on internet forums googled tips on how to evade taxes searched for an abortion clinic ordered a rainbow flag from an online store and done all sorts of other things that may not be illegal but your neighbors don’t necessarily need to know All this metadata about what you did online (and when you sent a WhatsApp message to whom and how often) has been stored at appropriately named Meta And you never worried about it Because you thought: “there’s so much data there I’m so uninteresting they’re really not going to catch me” But soon this argument will disappear Because profiles can soon be made even of all uninteresting people it was always just a matter of ‘enough computing power’
Oh they won’t do that
Such a profile of you isn’t just annoying if criminals get hold of it Profiles in government hands can also turn out quite unfavorable think for example: benefits scandal And that’s just ‘our own’ government All this data lies with American companies Subject to American laws It seems fair when you wonder if Americans are still our friends random citizens from other countries who weren’t on their friends list have sometimes ended up in Guantanamo Bay
Now such a gruesome scenario isn’t really expected yet But with enough computing power everyone can be followed where you are when and how long Continuously looking for a moment when you do something that can be labeled as ‘suspicious’ And anyone who potentially becomes slightly troublesome can soon even without pressing one button character assassination can be committed All data from everyone can selectively be bundled interpreted and presented as news fact And if the story doesn’t add up? No one finds out because content moderation no longer exists and through extreme-right populists concept ‘truth’ has long been hollowed out after which they filled this gap with unfounded general distrust And for those who don’t fear government it’s still naive to think American Big Tech would miss any opportunity to make extra money off you And there’s more money to be made off you if you’re angry and confused Chaos and division are lucrative
I’m not worried at all about an AI with self-awareness Super-rich creepy white men with far too much computing power: that’s what keeps me awake at night
Erwin Fisser 